trurl

I don't use the plexinc docker, I am using Linuxserver.io plex instead. But your plexinc docker mappings don't make sense to me. Why do you have 3 different container paths mapped to the same host path "/mnt/user/appdata/PlexMediaServer"? I just looked at the plexinc documentation on dockerhub, and your mappings don't seem to conform to those docs either. What is the purpose of the container path "/data" in your mappings? And wouldn't it make more sense if the "/transcode" container path was mapped to something other than the exact same path "/config" (your plex library) is mapped to so your transcodes aren't just dumped into that same folder with your plex library? None of that would cause filling docker image problem, but it does make me wonder if you know how to setup dockers. The reason I picked on NextCloud and Plex is because I am using those myself and have some idea what they are and how to set them up. One of the others you are running may be the cause of your problems but it would take a lot more research for me to figure out what you may have done wrong with those. Do you understand docker mappings of host paths to container paths? Do you know Linux is case sensitive? Do you know to use absolute and not relative paths to refer to the container paths within the application? Do you know why I'm asking these questions and what their importance is?

Are you sure you want that share to stay on cache? Why???

In other words, just setting it to cache-prefer is not enough. Mover can't move open files. While the docker service is enabled, mover can't move docker image since the docker service has it open. So it is unlikely you have actually gotten system share moved to cache. Post new diagnostics.

Way too difficult to read. If someone else gave something like that to you would you even try to make any sense of it? Maybe a screenshot would be better unless you want to go to the trouble to reformat it into clear columns. Really all I wanted is a list without all those details for each. A screenshot of the Docker section on the Dashboard page. I thought we could get into the other details as needed. And those details you dumped there will not really be sufficient for the problem of filling docker image since that also depends on settings within each application. Do you understand docker mappings of host paths to container paths? Do you know Linux is case sensitive? Do you know to use absolute and not relative paths to refer to the container paths within the application? Do you know why I'm asking these questions and what their importance is?

And what do you have for the transcode directory within the plex application? Do you use the plex DVR feature?

So have you tried it again after fixing that problem?

Are you using port multiplier or something? I would estimate 12TB parity sync to take about half that time.

Post docker run command for plex as explained at this very first link in the Docker FAQ: https://forums.unraid.net/topic/57181-docker-faq/?do=findComment&comment=564345

Is Nextcloud storing its files in the mariadb database or somewhere else?

stop array unassign disabled disk start array with disabled disk unassigned stop array reassign disabled disk start array to begin rebuild

Working for me

20GB docker image should be more than enough. You almost certainly have some application configured so it is writing to a path that isn't mapped. Which dockers do you run? Go to Tools - Diagnostics and attach the complete diagnostics ZIP file to your NEXT post.

Disk looks OK. Check connections. Do you know how to rebuild to the same disk?

I've seen these leftover system share folders/files on the array even when the user is supposedly following my step-by-step instructions which included disabling the services. It looked to me at least in some cases that they were not moved precisely because they were duplicates, since the user would have system share on both cache and array before we started to make any corrections. Don't know how they came to be created as duplicates though. No doubt they were created on the array since cache wasn't installed at the time the services were first enabled, but I don't know how they came to have new ones created on cache later. And I have seen it where just the folder is all that still exists on the array, and there are no files in it. Or it may be that it is usually the case that only the folder still exists. We have always just finished the cleanup manually.

Why didn't you ask for help on that support thread you linked instead of starting a new topic?

Delete them from disk1. Also, the link you gave for CA Backup above was for the deprecated version (as it says in the title). You should use CA Backup V2. You can install CA Backup V2 from the Apps page. You should always go from the Apps page to install anything so you only get current and compatible addons.

You didn't actually mention if you had tried the ? Help button suggested. That Help button is a toggle, that turns Help on/off for the whole webUI. You can also toggle help for a specific setting on/off by clicking on the label for the setting.

Did you enter any credentials at all? If the current Windows username and password will work to access the share, then it won't even ask you to login.

My guess is you have one or more of your dockers misconfigured. One reason I say that is because you have given docker image 50G, when 20G should be more than enough, and you have already used 23G of the 50G you have allocated. This most likely means you have an application writing to a path that isn't mapped. An unmapped path is a path inside the docker image. However, that won't actually cause you to run out of RAM. It will just cause your docker image to fill up. Making it larger than 20G will not fix that problem, it will just make it take longer to fill. What will fill RAM though is a docker mapping that isn't to actual storage. Any host path that isn't to the user shares or the disks is a path in RAM. So, you probably have both situations. Also, your appdata has files on the array even though it is cache-prefer. Not sure how you got them there unless you had changed the cache setting for appdata, or else you filled cache and appdata had to overflow to the array. Those probably can't be moved back to cache while dockers are running because mover can't move open files. We will probably have to tear down your dockers and figure out what you have done with them. Which dockers do you run?

Yes, going by IP vs servername makes Windows treat it as a different network computer and so different credentials can be used. Even though Windows will sometimes prompt you for a different login because permission is denied, it won't actually use whatever you enter at the prompt since you are already logged in with different credentials. Definitely a Windows quirk, since it treats other systems (even other Windows on the network) in just the same way.

CRC errors mean the data was corrupted between the disk and the rest of the computer. Usually a bad connection.

Syslogs are spammed with May 8 12:43:29 juggernaut root: error: /plugins/unassigned.devices/UnassignedDevices.php: wrong csrf_token

Go to Tools - Diagnostics and attach the complete Diagnostics ZIP file to your NEXT post.

If you put music1 and music2 as subfolders under music, then music1 and music2 become folders within the music share. and music1 and music2 will no longer be user shares themselves. Don't know if that is what you intend or not.

I see your other thread. Looks like you knew how to rebuild a disk. He hasn't even gotten that far so he isn't having the same problems you are at all.